arXiv subjects

Shen Xu

Publications and source records attributed to Shen Xu.

8 recordsLinked to original sources

arXiv2026

Representing Entity Importance in AI Knowledge Systems: A Dual-Signal Framework of Audience Evaluation and Structural Authority

AI knowledge systems require representations of entity importance for retrieval, recommendation, evidence selection, and knowledge-intensive reasoning. Yet importance is often reduced to a single score derived from either human response or graph structure. Such compression may discard distinctions that matter when an AI system must choose among entities for different tasks. This study introduces an interpretable dual-signal representation in which each entity is characterized by an audience-evaluation dimension and a structural-authority dimension. The framework is evaluated using movie entities as an empirical validation domain. IMDb non-commercial datasets provide a rating-based audience ranking, Wikidata supports entity alignment, and English Wikipedia hyperlinks form the knowledge network on which PageRank estimates structural authority. Experiments on 482 entities and 13,690 directed relationships reveal a statistically significant but weak association between the two dimensions (Spearman rho = 0.2275, p < 0.001). Their overlap is only 10% in the top 10 and 34% in the top 100, while entity-level divergence occurs in both directions. The results show that audience evaluation and structural authority are non-redundant signals and should not automatically be collapsed into a single scalar notion of importance. The contribution is not a new ranking algorithm or learned embedding, but a minimal knowledge-representation framework and an empirical test of its dimensional necessity. The findings support task-aware AI knowledge systems that preserve distinct importance signals before applying context-specific selection or aggregation.

cs.AI↗

arXiv2026

HCInfer: An Efficient Inference System via Error Compensation for Resource-Constrained Devices

LLMs often struggle with memory-constrained deployment on consumer-grade hardware due to their massive parameter sizes. While existing solutions such as model compression and offloading improve deployment feasibility, they often suffer from substantial accuracy degradation or severe throughput bottlenecks. Recent error compensation methods recover accuracy through auxiliary LoRA-style branches, and we observe that these branches are inherently amenable to offloading: they require substantial parameter storage but access only a small subset of compensation parameters during each inference step. Motivated by this opportunity, we propose HCInfer, a heterogeneous inference system that offloads residual compensation to the CPU while executing the compressed backbone on the GPU, and further introduces an asynchronous compensation pipeline and sensitivity-aware dynamic rank allocation to hide compensation overhead and maximize accuracy recovery. Experimental results show that HCInfer achieves a maximum accuracy improvement of 5.2% on downstream tasks compared to compression model and sustaining a maximum speedup of 10.4x compared to full-precision model.

cs.LG↗

arXiv2023

An ontology-aided, natural language-based approach for multi-constraint BIM model querying

Being able to efficiently retrieve the required building information is critical for construction project stakeholders to carry out their engineering and management activities. Natural language interface (NLI) systems are emerging as a time and cost-effective way to query Building Information Models (BIMs). However, the existing methods cannot logically combine different constraints to perform fine-grained queries, dampening the usability of natural language (NL)-based BIM queries. This paper presents a novel ontology-aided semantic parser to automatically map natural language queries (NLQs) that contain different attribute and relational constraints into computer-readable codes for querying complex BIM models. First, a modular ontology was developed to represent NL expressions of Industry Foundation Classes (IFC) concepts and relationships, and was then populated with entities from target BIM models to assimilate project-specific information. Hereafter, the ontology-aided semantic parser progressively extracts concepts, relationships, and value restrictions from NLQs to fully identify constraint conditions, resulting in standard SPARQL queries with reasoning rules to successfully retrieve IFC-based BIM models. The approach was evaluated based on 225 NLQs collected from BIM users, with a 91% accuracy rate. Finally, a case study about the design-checking of a real-world residential building demonstrates the practical value of the proposed approach in the construction industry.

cs.CL↗

arXiv2022

Ballistic Thermal Transport at Sub-10 nm Laser-Induced Hot Spots in GaN Crystal

Gallium nitride (GaN) is a typical wide-bandgap semiconductor with a critical role in a wide range of electronic applications. Ballistic thermal transport at nanoscale hotspots will greatly reduce the performance of a device when its characteristic length reaches the nanometer scale, due to heat dissipation. In this work, we developed a tip-enhanced Raman thermometry approach to study ballistic thermal transport within the range of 10 nm in GaN, simultaneously achieving laser heating and measuring the local temperature. The Raman results showed that the temperature increase from an Au-coated tip-focused hotspot was up to two times higher (40 K) than that in a bare tip-focused region (20 K). To further investigate the possible mechanisms behind this temperature difference, we performed electromagnetic simulations to generate a highly focused heating field, and observed a highly localized optical penetration, within a range of 10 nm. The phonon mean free path (MFP) of the GaN substrate could thus be determined by comparing the numerical simulation results with the experimentally measured temperature increase which was in good agreement with the average MFP weighted by the mode-specific thermal conductivity, as calculated from first-principles simulations. Our results demonstrate that the phonon MFP of a material can be rapidly predicted through a combination of experiments and simulations, which can find wide application in the thermal management of GaN-based electronics.

cond-mat.mes-hall↗

arXiv2022

Ballistic Transport Enhanced Heat Convection at Nanoscale Hotspots

Along with device miniaturization, severe heat accumulation at unexpected nanoscale hotspots attracts wide attentions and urges efficient thermal management. Heat convection is one of the important heat dissipating paths but its mechanism at nanoscale hotspots is still unclear. Here shows the first experimental investigation of the convective heat transfer coefficient at size-controllable nanoscale hotspots. A specially designed structure of a single layer graphene supported by gold nanorods (AuNRs) is proposed, in which the AuNRs generate plasmonic heating sources of the order of hundreds of nanometers under laser irradiation and the graphene layer works as a temperature probe in Raman thermometry. The determined convective heat transfer coefficient is found to be about three orders of magnitude higher than that of nature convection, when the simultaneous interfacial heat conduction and radiation are carefully evaluated. Heat convection thus accounts to more than half of the total energy transferred across the graphene/AuNRs interface. Both the plasmonic heating induced nanoscale hotspots and ballistic convection of gas molecules contribute to the enhanced heat convection. This work reveals the importance of heat convection at nanoscale hotspots to the accurate thermal design of miniaturized electronics, and further offers a new way to evaluate the convective heat transfer coefficient at nanoscale hotspots.

physics.app-ph↗

arXiv2016

Strongly Anisotropic Thermal and Electrical Conductivities of Self-assembled Silver Nanowire Network

Heat dissipation issues are the emerging challenges in the field of flexible electronics. Thermal management of flexible electronics creates a demand for flexible materials with highly anisotropic thermal conductivity, which work as heat spreaders to remove excess heat in the in-plane direction and as heat shields to protect human skin or device components under them from heating. This study proposes a self-assembled silver nanowire network with high thermal and electrical anisotropy with the potential to solve these challenges. The in-plane thermal conductivity of the network along the axial direction of silver nanowires is measured as 37 W/m-K while the cross-plane thermal conductivity is only 0.36 W/m-K. The results of measurements of electrical and thermal conductivities suggest that abundant wire-wire contacts strongly impede thermal transport. The excellent alignment of nanowires results in the same anisotropy ratio of 3 for both thermal and electrical conduction in the two in-plane directions. The ratio remains unchanged as the temperature decrease to 50 K, which indicates that wire-wire contacts lower the thermal and electrical conduction in the two directions to the same extent and their effect is independent of temperature. In addition, phonon softening markedly reduces the Debye temperatures of the network, which are fitted from the electrical resistivity data. As a result of phonon thermal conduction, the Lorenz numbers of the film in the two directions, which are approximately the same, are larger than the Sommerfeld value at room temperature and decrease as temperature decreases because of small angle scattering and the reduced phonon contribution. This nanowire network provides a solution to the emerging challenges of thermal management of flexible electronics.

cond-mat.mes-hall↗

arXiv2014

Phonon Softening and Weak Temperature-dependent Lorenz Number for Bio-supported Ultra-thin Ir Film

This work reports on the first-time study of the temperature-dependent behavior of the Lorenz number of bio-supported average 3.2 nm-thin Ir film down to 10 K. Due to the strong imperfection-electron scattering, a very large residual resistivity is observed for the film that dominates the overall electron transport. The Debye temperature (221 K)of the film is found much smaller than that of bulk (308 K). This phonon softening strongly confirms the extensive surface and grain boundary electron scatterings. More than one order of magnitude reduction is observed for the thermal conductivity of the film. We find the Wiedemann-Franz Law still applies to our film even at low temperatures. The overall Lorenz number and that of imperfection structure are close to the Sommerfeld value and shows little temperature dependence. This is contrast to other studied low dimensional metallic structures that have a much larger Lorenz number. Electron tunneling and hopping in the biomaterial substrate are speculated responsible for the observed Lorenz number.

cond-mat.mes-hall↗

arXiv2014

Temperature Dependent Behavior of Thermal Conductivity of Sub-5 nm Ir film: Defect-electron Scattering Quantified by Residual Thermal Resistivity

By studying the temperature-dependent behavior of electron thermal conductivity (k) in a 3.2 nm-thin film, we quantify the extremely confined defect-electron scattering and reveal the intrinsic phonon-electron scattering that is shared by bulk Ir. At low temperatures below 50 K, the thermal conductivity of the thin film has almost two orders of magnitude reduction from that of the bulk Ir. The thermal conductivity of the film increases with increasing temperature while that of the bulk Ir has an opposite trend. We introduce a unified thermal resistivity to interpret this completely different k-T relation. This residual thermal resistivity provides an unprecedented way to quantitatively evaluating defect-electron scatterings in heat conduction. The interfacial thermal conductance across the grain boundaries is found larger than that of the Al/Cu interface. Its value is proportional to temperature largely because of the electron's specific heat. A unified interfacial thermal conductance is defined and firmly proves such relation. The electron reflection coefficient is found to be large (88%) and almost temperature independent. This means most of the electrons which scatter with the grain boundary would be reflected back and the scatterings are not affected by temperature.

cond-mat.mes-hall↗